If the genetics you have used do not allow for feather sexing, you will have to resort to vent sexing. Care must be taken in the feather sexing to ensure you are not calling some males female due to differences in feather length. Slower feather growth in the lower layer of wing primaries in males sometimes makes them look female. If you check where the feathers enter the wing you will be able to see if it is the upper or lower layer of feathers that is actually shorter. When training your staff or students, it is valuable to do some quality assurance checks on your females after sorting and show the anomalies to the group for teaching purposes.
Vent sexing is difficult to do without proper training. I have worked beside chick sexers and know what I'm supposed to be looking for, but can't do it reliably myself. We hire experts to sex our older genetic stocks that can't be feather-sexed.
I agree with Dr Renema, feather sexing is much easier if you considered good training . But actually we have a project to conserve and characterize local chicken genetic lines, through this project sexing is a very important issue. So, i sent one of the MSc candidates in my team to get trained on vent sexing, it takes about 10 training sessions each of 5 hours. The output was great he is very good at vent sexing which is highly supported our project. In addition, currently he is training some of his colleagues too. So if you able to train somebody of your team on vent sexing it would be great option.
Use a breed (such as Lohmann Brown) where colour is sex linked-it's easy then. Otherwise it will take time to get someone trained using vent sexing. It's the best way.
The autosexing by using sex linked traits like slow and rapid feathering , represent the best method , but this method need control the mating of the parents , the female must be slow feathering ( the dominant allele ) and the male be rapid feathering ( recessive allele ) , in this case the slow feather chicks will be males according to crisscross inheritance .